Portfolio of Leveraged Exchange Traded Funds

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
Beliavsky
Posts: 932
Joined: Sun Jun 29, 2014 10:21 am

Portfolio of Leveraged Exchange Traded Funds

Post by Beliavsky » Wed Jan 09, 2019 4:13 pm

The idea of this working paper is that by using leveraged ETFs for some asset classes, one can create more space in the portfolio to own BND (Vanguard Total Bond Market ETF), which historically has had a high Sharpe ratio.

A Portfolio of Leveraged Exchange Traded Funds
William Trainor
East Tennessee State University
Indudeep Chhachhi
Western Kentucky University
Chris Brown
Western Kentucky University
Date Written: October 24, 2018
Abstract
Leveraged exchange traded funds (LETFs) are marketed as short-term trading vehicles that magnify the daily returns of an underlying index. With the proliferation of LETFs over the last 10 years, a diversified portfolio that mimics the returns of a 100% investment can be created using only a fraction of the investor's wealth. Results suggest a portfolio created with LETFs outperforms a portfolio using traditional ETFs by approximately 0.6% to 1.4% annually by investing the excess wealth in a diversified or short to mid-duration bond portfolio. Downside risk is reduced using LETFs because the majority of the LETF portfolio is invested in a relatively safe bond fund.

Keywords: Diversified Portfolios, Leveraged Exchange Traded Funds

User avatar
danielc
Posts: 640
Joined: Sun Dec 10, 2017 4:48 am
Location: Iowa, USA
Contact:

Re: Portfolio of Leveraged Exchange Traded Funds

Post by danielc » Wed Jan 09, 2019 4:30 pm

Beliavsky wrote:
Wed Jan 09, 2019 4:13 pm
Leveraged exchange traded funds (LETFs) are marketed as short-term trading vehicles that magnify the daily returns of an underlying index.
Short-term trading is gambling, not investing. It doesn't make sense to write than sentence, and then talk about 10 year returns, or reducing downside risk.

Topic Author
Beliavsky
Posts: 932
Joined: Sun Jun 29, 2014 10:21 am

Re: Portfolio of Leveraged Exchange Traded Funds

Post by Beliavsky » Wed Jan 09, 2019 4:33 pm

danielc wrote:
Wed Jan 09, 2019 4:30 pm
Beliavsky wrote:
Wed Jan 09, 2019 4:13 pm
Leveraged exchange traded funds (LETFs) are marketed as short-term trading vehicles that magnify the daily returns of an underlying index.
Short-term trading is gambling, not investing. It doesn't make sense to write than sentence, and then talk about 10 year returns, or reducing downside risk.
The strategies studied in the paper require only monthly trading of leveraged ETFs for rebalancing. Rebalancing is commonly advocated by Bogleheads.

User avatar
danielc
Posts: 640
Joined: Sun Dec 10, 2017 4:48 am
Location: Iowa, USA
Contact:

Re: Portfolio of Leveraged Exchange Traded Funds

Post by danielc » Wed Jan 09, 2019 4:48 pm

Beliavsky wrote:
Wed Jan 09, 2019 4:33 pm
The strategies studied in the paper require only monthly trading of leveraged ETFs for rebalancing. Rebalancing is commonly advocated by Bogleheads.
MONTHLY rebalancing is never advocated by Bogleheads. Bogleheads argue about whether yearly rebalancing is too frequent. A lot of us think that every 2-3 years is the right frequency. And we don't use leverage.

Starfish
Posts: 1461
Joined: Wed Aug 15, 2018 6:33 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by Starfish » Wed Jan 09, 2019 5:03 pm

I disagree. Most people leverage. For example everybody having loans and investments at the same time.

alex_686
Posts: 5097
Joined: Mon Feb 09, 2015 2:39 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by alex_686 » Wed Jan 09, 2019 5:08 pm

danielc wrote:
Wed Jan 09, 2019 4:30 pm
Beliavsky wrote:
Wed Jan 09, 2019 4:13 pm
Leveraged exchange traded funds (LETFs) are marketed as short-term trading vehicles that magnify the daily returns of an underlying index.
Short-term trading is gambling, not investing. It doesn't make sense to write than sentence, and then talk about 10 year returns, or reducing downside risk.
It is not gambling nor rebalancing. It is about creating a synthetic exposure to equities. It is not exactly Bogleheads because it is complex. This does not mean it is wrong or that it is a active strategy.

User avatar
Phineas J. Whoopee
Posts: 8819
Joined: Sun Dec 18, 2011 6:18 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by Phineas J. Whoopee » Wed Jan 09, 2019 5:12 pm

Starfish wrote:
Wed Jan 09, 2019 5:03 pm
I disagree. Most people leverage. For example everybody having loans and investments at the same time.
The question is not about leverage in general. It's about the specific workings of leveraged and inverse ETFs, which do not operate the way one might expect based only on their names.

From time to time we have threads where people claim to be immune from the built-in decay over time from continually holding such securities well beyond their design lengths.

I don't know why.

I don't deny one can backtest specific scenarios where one would have made money doing that. I'm sure there must be some.

PJW

User avatar
danielc
Posts: 640
Joined: Sun Dec 10, 2017 4:48 am
Location: Iowa, USA
Contact:

Re: Portfolio of Leveraged Exchange Traded Funds

Post by danielc » Wed Jan 09, 2019 5:15 pm

alex_686 wrote:
Wed Jan 09, 2019 5:08 pm
It is not gambling nor rebalancing. It is about creating a synthetic exposure to equities. It is not exactly Bogleheads because it is complex. This does not mean it is wrong or that it is a active strategy.
I didn't say that it is wrong. People can do whatever they want with their money. I think monthly rebalancing leveraged ETFs is both active and risky. I don't understand how trading every month can be considered anything but "active".

ThrustVectoring
Posts: 769
Joined: Wed Jul 12, 2017 2:51 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by ThrustVectoring » Wed Jan 09, 2019 5:19 pm

Leverage is leverage, and levered ETFs aren't the cheapest source of leverage. Treasury futures are. There's a small advantage to the levered ETFs due to convexity - as the stock market declines, the daily rebalance will reduce your position automatically, limiting extreme downside losses to no more than the initial investment. But overall, if you want convexity, you can simply buy it via index options, and likely at a better price to boot.
Current portfolio: 60% VTI / 40% VXUS

User avatar
Phineas J. Whoopee
Posts: 8819
Joined: Sun Dec 18, 2011 6:18 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by Phineas J. Whoopee » Wed Jan 09, 2019 5:30 pm

ThrustVectoring wrote:
Wed Jan 09, 2019 5:19 pm
Leverage is leverage, and levered ETFs aren't the cheapest source of leverage. ...
A daily, or even monthly, double return ETF is nothing like borrowing to buy an index ETF. Leverage is not all the same.

PJW

EfficientInvestor
Posts: 237
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: Portfolio of Leveraged Exchange Traded Funds

Post by EfficientInvestor » Wed Jan 09, 2019 5:46 pm

After skimming through the paper, my main takeaway is that their asset allocations need to be revisited. They are basically taking a 70/30 portfolio, applying leverage to it, and then backing down risk by allocating funds to a relatively risk-free return investment like BND (VBMFX). The problem is that leverage is being applied to an allocation that is inefficient to begin with. Instead, they should flip it around and have something closer to a 30/70 portfolio and then apply leverage. The backtest below uses 2X stock funds that have been around since 2000 and uses WHOSX as a proxy for a 2X bond fund somewhere between 7 and 20 years in duration. Portolio 1 represents the 2X allocations presented in Table 2 of the paper. Portfolio 2 are my proposed updates. Portfolio 3 has my proposed allocations with risk reduced via VBMFX. Portfolio 3 results in the same annualized returns as Portfolio 1, but with 1/3 the max drawdown.

https://www.portfoliovisualizer.com/bac ... tion5_3=50

Mar 2000 - Dec 2018
Port 1 - CAGR = 6.5%, Max DD = -30.0
Port 2 - CAGR = 8.3%, Max DD = -20.7%
Port 3 - CAGR = 6.6%, Max DD = -11.4%
S&P 500 - CAGR = 5.2%, Max DD = -51.0%

alex_686
Posts: 5097
Joined: Mon Feb 09, 2015 2:39 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by alex_686 » Wed Jan 09, 2019 6:25 pm

danielc wrote:
Wed Jan 09, 2019 5:15 pm
alex_686 wrote:
Wed Jan 09, 2019 5:08 pm
It is not gambling nor rebalancing. It is about creating a synthetic exposure to equities. It is not exactly Bogleheads because it is complex. This does not mean it is wrong or that it is a active strategy.
I didn't say that it is wrong. People can do whatever they want with their money. I think monthly rebalancing leveraged ETFs is both active and risky. I don't understand how trading every month can be considered anything but "active".
Some people get their equity exposure via futures. Each month they roll over to a new set of contracts. Is this active? Rules based, broad exposure to the market (a.k.a. Beta), no active choices done in weighting, specific securities. It is trying to replicate and reflect the market. So yes, this is passive.

ThrustVectoring
Posts: 769
Joined: Wed Jul 12, 2017 2:51 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by ThrustVectoring » Wed Jan 09, 2019 6:29 pm

Phineas J. Whoopee wrote:
Wed Jan 09, 2019 5:30 pm
ThrustVectoring wrote:
Wed Jan 09, 2019 5:19 pm
Leverage is leverage, and levered ETFs aren't the cheapest source of leverage. ...
A daily, or even monthly, double return ETF is nothing like borrowing to buy an index ETF. Leverage is not all the same.

PJW
I'm well aware of the buy high / sell low effect of periodic rebalancing of short or levered positions. The effect you've listed is simply the natural consequence of daily rebalancing a levered product; when a levered position loses value, the underlying securities don't lose as much value, so you need to sell in order to maintain your leverage.

If you're borrowing to buy an index ETF, you're implementing a different strategy, so of course you'll wind up with different results. This isn't attributable to the leverage you're using being magically different by virtue of being stuffed inside an ETF. If you behaved like the ETF does, you'd get their results. The fundamental problem with the levered ETF isn't the leverage, it's that their strategy is dumb for long-term investors due to using more aggressive risk-taking than the Kelley optimum, guaranteeing that it will lose money over time due to buying high and selling low to maintain their desired leverage.
Current portfolio: 60% VTI / 40% VXUS

User avatar
hdas
Posts: 1271
Joined: Thu Jun 11, 2015 8:24 am

Re: Portfolio of Leveraged Exchange Traded Funds

Post by hdas » Wed Jan 09, 2019 7:47 pm

This is the poor man’s version of portable alpha. Nothing wrong with it. Except the high cost and the effects of daily rebalancing. One suggests to wait til your notional is big enough to implement with futures.

PS- Perhaps even more interesting is the implementation with options.
"whenever there is a randomized way of doing something, then there is a nonrandomized way that delivers better performance but requires more thought" ET Jaynes

User avatar
danielc
Posts: 640
Joined: Sun Dec 10, 2017 4:48 am
Location: Iowa, USA
Contact:

Re: Portfolio of Leveraged Exchange Traded Funds

Post by danielc » Wed Jan 09, 2019 9:11 pm

alex_686 wrote:
Wed Jan 09, 2019 6:25 pm
Some people get their equity exposure via futures. Each month they roll over to a new set of contracts. Is this active?
I would call that active.
alex_686 wrote:
Wed Jan 09, 2019 6:25 pm
Rules based, broad exposure to the market (a.k.a. Beta), no active choices done in weighting, specific securities. It is trying to replicate and reflect the market. So yes, this is passive.
I don't equate "rules based" with "passive". Most active investors use rules. Most day traders use rules. Heck, if you write a computer program to do day trading for you, that's obviously based on rules. None of that changes the fact that day trading is somewhere between gambling and speculation.

User avatar
hdas
Posts: 1271
Joined: Thu Jun 11, 2015 8:24 am

Re: Portfolio of Leveraged Exchange Traded Funds

Post by hdas » Thu Jan 10, 2019 1:05 am

danielc wrote:
Wed Jan 09, 2019 9:11 pm
alex_686 wrote:
Wed Jan 09, 2019 6:25 pm
Some people get their equity exposure via futures. Each month they roll over to a new set of contracts. Is this active?
I would call that active.
There’s nothing active about that if you are just rolling your futures to maintain exposure. It’s as passive as it gets. Cheers :greedy
"whenever there is a randomized way of doing something, then there is a nonrandomized way that delivers better performance but requires more thought" ET Jaynes

long_gamma
Posts: 352
Joined: Mon Sep 14, 2015 4:13 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by long_gamma » Thu Jan 10, 2019 7:15 am

danielc wrote:
Wed Jan 09, 2019 9:11 pm
alex_686 wrote:
Wed Jan 09, 2019 6:25 pm
Some people get their equity exposure via futures. Each month they roll over to a new set of contracts. Is this active?
I would call that active.
If this is your definition then VFINX is an active fund.

https://personal.vanguard.com/pub/Pdf/p040.pdf
From their prospectus
"To help stay fully invested and to reduce transaction costs, the Fund
may invest, to a limited extent, in derivatives, including equity futures. The Fund may
also use derivatives such as total return swaps to obtain exposure to a stock, a basket of
stocks, or an index. Generally speaking, a derivative is a financial contract whose value
is based on the value of a financial asset (such as a stock, a bond, or a currency), a
physical asset (such as gold, oil, or wheat), a market index (such as the S&P 500 Index),
or a reference rate (such as LIBOR). Investments in derivatives may subject the Fund to
risks different from, and possibly greater than, those of investments directly in the
underlying securities or assets. "
"Everyone has a plan 'till they get punched in the mouth." --Mike Tyson

User avatar
danielc
Posts: 640
Joined: Sun Dec 10, 2017 4:48 am
Location: Iowa, USA
Contact:

Re: Portfolio of Leveraged Exchange Traded Funds

Post by danielc » Thu Jan 10, 2019 8:52 am

hdas wrote:
Thu Jan 10, 2019 1:05 am
danielc wrote:
Wed Jan 09, 2019 9:11 pm
alex_686 wrote:
Wed Jan 09, 2019 6:25 pm
Some people get their equity exposure via futures. Each month they roll over to a new set of contracts. Is this active?
I would call that active.
There’s nothing active about that if you are just rolling your futures to maintain exposure. It’s as passive as it gets. Cheers :greedy
We are literally talking about monthly trading. How is that not active?

User avatar
danielc
Posts: 640
Joined: Sun Dec 10, 2017 4:48 am
Location: Iowa, USA
Contact:

Re: Portfolio of Leveraged Exchange Traded Funds

Post by danielc » Thu Jan 10, 2019 8:54 am

long_gamma wrote:
Thu Jan 10, 2019 7:15 am
danielc wrote:
Wed Jan 09, 2019 9:11 pm
alex_686 wrote:
Wed Jan 09, 2019 6:25 pm
Some people get their equity exposure via futures. Each month they roll over to a new set of contracts. Is this active?
I would call that active.
If this is your definition then VFINX is an active fund.

https://personal.vanguard.com/pub/Pdf/p040.pdf
From their prospectus
"To help stay fully invested and to reduce transaction costs, the Fund
may invest, to a limited extent, in derivatives, including equity futures. The Fund may
also use derivatives such as total return swaps to obtain exposure to a stock, a basket of
stocks, or an index. Generally speaking, a derivative is a financial contract whose value
is based on the value of a financial asset (such as a stock, a bond, or a currency), a
physical asset (such as gold, oil, or wheat), a market index (such as the S&P 500 Index),
or a reference rate (such as LIBOR). Investments in derivatives may subject the Fund to
risks different from, and possibly greater than, those of investments directly in the
underlying securities or assets. "
What percentage of VFINX is in derivatives? The paragraph you quoted says "to a limited extent".

foo.c
Posts: 62
Joined: Sat Jul 15, 2017 4:55 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by foo.c » Thu Jan 10, 2019 9:10 am

This is similar to what I did last year except I did more trading and had some hedges. It worked great in January. Just about returned to my high water mark in November. Lost all my gains in December. I finished up 0.6%, which is still better than the market did, but wasn't worth the effort. I'm kind of in a re-evaluation of things at the moment.

This is in a small account I use to keep myself occupied so I don't mess with my real investments which are vanilla 3 fund passive.

User avatar
nisiprius
Advisory Board
Posts: 39463
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Portfolio of Leveraged Exchange Traded Funds

Post by nisiprius » Thu Jan 10, 2019 9:45 am

Let's try a quick real-world reality check. Here's what I'm going to do. I'm going to create a 60/40 portfolio using the Vanguard S&P 500 index fund, VFINX, for stocks, and Total Bond, VBMFX, for bonds. Then I'm going to create a second portfolio using ULPIX, the ProFunds Ultrabull (2X S&P 500) for stocks instead; I'm going to select "monthly rebalancing" in PortfolioVisualizer; and I'm going to hand-adjust the allocation until I get as close as I can to the same standard deviation as I do for the "straight" 60/40.

This is a real strategy that could have been done in the real world using real funds, and it gives us over twenty years of data including about two market cycles. The specific fund is also one that was called out by name by two Yale professors in a book urging that young investors use a 200% stock allocation in their early years, as being one suitable way to get the 2X leverage, so it has an expert endorsement of sorts.

I have not done this yet. I haven't looked at it. I'm going to do it and then I am going to post the results, however they turn out.

Portfolio 1, blue, is the conventional 60/40 portfolio using unleveraged mutual funds.
Portfolio 2, red, is the portfolio that substitutes a 2X leveraged mutual fund for stocks, enabling us to boost the bond allocation.

Source

Image

Image

I think the fairest way to describe this is that...

in real life, no advantage would have been seen in using a leveraged fund portfolio (red line) over a conventional fund portfolio.

The differences are small, but in each case the advantage was with the conventional portfolio. It had a higher final value, a higher CAGR (0.52% higher, annualized), a slightly better best year, a slightly less bad worse year, a slightly lower drawdown in 2008-2009, a lower Sharpe ratio, and a lower Sortino ratio.

Nothing terrible would have happened to the investor using the leveraged fund.

I think this is important. The authors of the paper make a very appealing claim in their abstract:
Downside risk is reduced using LETFs because the majority of the LETF portfolio is invested in a relatively safe bond fund.
This is intuitively appealing, but in the real world--despite increasing the bond allocation from 40% in the conventional portfolio to 70% in the leveraged portfolio...

no reduction in downside risk would have been seen in in the leveraged portfolio, either in 2008-2009 or in 2000-2002.
Last edited by nisiprius on Thu Jan 10, 2019 10:05 am, edited 1 time in total.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

User avatar
nisiprius
Advisory Board
Posts: 39463
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Portfolio of Leveraged Exchange Traded Funds

Post by nisiprius » Thu Jan 10, 2019 10:04 am

If we repeat the same exercise, using the same company's 2X S&P 500 ETF, SSO--which restricts us to using only about half as long a time period, 2007-present--we see that there would no longer have been a visible disadvantage, but neither would there have been any advantage. No pain, no gain. Plus, of course, you need to rebalance monthly, somehow, which in the real world is not trivial to do.

Portfolio 1, blue, is 60/40 VFINX/VBMFX; portfolio 2, red, is 29.0% ULPIX, 71.0% VBMFX.

Again, I think it is noteworthy that the authors' claim that "downside risk is reduced" was not actually borne out here, despite the use of a much larger bond allocation in portfolio 2 (red).

Source

Image
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

EfficientInvestor
Posts: 237
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: Portfolio of Leveraged Exchange Traded Funds

Post by EfficientInvestor » Thu Jan 10, 2019 10:37 am

nisiprius wrote:
Thu Jan 10, 2019 10:04 am
If we repeat the same exercise, using the same company's 2X S&P 500 ETF, SSO--which restricts us to using only about half as long a time period, 2007-present--we see that there would no longer have been a visible disadvantage, but neither would there have been any advantage. No pain, no gain. Plus, of course, you need to rebalance monthly, somehow, which in the real world is not trivial to do.

Portfolio 1, blue, is 60/40 VFINX/VBMFX; portfolio 2, red, is 29.0% ULPIX, 71.0% VBMFX.

Again, I think it is noteworthy that the authors' claim that "downside risk is reduced" was not actually borne out here, despite the use of a much larger bond allocation in portfolio 2 (red).
(Updated to include DXKLX for 2X 7-10 treasury fund)
I don't believe the flaw is in their use of leveraged funds. As stated in my earlier post, I believe the flaw is in how they apply leverage. They are applying leverage to an inefficient 70/30 asset allocation and then ramping down risk by transferring allocation to a relatively risk-free bond fund like VBMFX/BND. What they should have done is set the allocations to an efficient 30/70 and then apply leverage. The backtest below shows that leverage applied to a more efficient (high Sharpe ratio) allocation can be quite beneficial. The backtest uses SPY (S&P 500) and IEF (7-10 year treasuries) for the unleveraged fund and SSO (2X S&P 500) and DXKLX (2X 7-10 year treasury) for leveraged funds. See comparison of these funds at second link below.

https://www.portfoliovisualizer.com/bac ... tion5_3=40

Portfolio 1 is a 60/40 portfolio. Portfolio 2 (2X 30/70) would have had higher CAGR than a 100/0 portfolio over the time period with less drawdown than a 60/40 portfolio. Then, after you have applied leverage to the 30/70 portfolio, you can adjust risk down using VBMFX. For instance, Portfolio 3 reduces risk by moving 50% allocation to VBMFX and splits the remaining 60% at a 30/70 ratio between the 2X stock and bond funds. This portfolio would have achieved the same return as 60/40 with less than half the max drawdown.

Jul 2006 - Dec 2018
P1 - CAGR = 7.3%, Max DD = -26.8%
P2 - CAGR = 10.5%, Max DD = -17.9%
P3 - CAGR = 7.3%, Max DD = -10.8%
S&P 500 - CAGR = 7.7%, Max DD = -51.0%

Similar relative results if we go further back in time with slightly different funds. WHOSX is used as proxy for 2X 7-10 year treasury. See comparison at second link below.

Dec 1997 - Dec 2018
P1 - CAGR = 6.4%, Max DD = -28.0%
P2 - CAGR = 8.3%, Max DD = -20.6%
P3 - CAGR = 6.6%, Max DD = -10.9%
S&P 500 - CAGR = 6.6%, Max DD = -51.0%

https://www.portfoliovisualizer.com/bac ... tion5_3=50

WHOSX vs DKXLX: https://www.portfoliovisualizer.com/bac ... ion2_2=100

Image
Last edited by EfficientInvestor on Thu Jan 10, 2019 11:50 am, edited 2 times in total.

long_gamma
Posts: 352
Joined: Mon Sep 14, 2015 4:13 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by long_gamma » Thu Jan 10, 2019 11:20 am

nisiprius wrote:
Thu Jan 10, 2019 10:04 am
If we repeat the same exercise, using the same company's 2X S&P 500 ETF, SSO--which restricts us to using only about half as long a time period, 2007-present--we see that there would no longer have been a visible disadvantage, but neither would there have been any advantage. No pain, no gain. Plus, of course, you need to rebalance monthly, somehow, which in the real world is not trivial to do.

Portfolio 1, blue, is 60/40 VFINX/VBMFX; portfolio 2, red, is 29.0% ULPIX, 71.0% VBMFX.

Again, I think it is noteworthy that the authors' claim that "downside risk is reduced" was not actually borne out here, despite the use of a much larger bond allocation in portfolio 2 (red).
This analysis is flawed. You need to lever up both equity and bonds. I used same process as yours, but levered up bonds also. To control the standard deviation and bring it to same as 60/40, I will use short term bonds.

Portfolio 1:
SSO 30%, DXKLX (2 times treasury bull) 39%, SHY 31%
Portfolio 2: VFINX 60%, and VBMFX 40%.

It is even balanced annually so no extra cost is involved.

https://www.portfoliovisualizer.com/ba ... tion5_1=31

Image
"Everyone has a plan 'till they get punched in the mouth." --Mike Tyson

garlandwhizzer
Posts: 2512
Joined: Fri Aug 06, 2010 3:42 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by garlandwhizzer » Thu Jan 10, 2019 12:33 pm

It is a very simple exercise in mathematics over a given time frame to create a complex portfolio using leverage/alternates/options, etc., with higher bond allocations that outperforms a standard 60/40 portfolio and provides a higher Sharpe ratio. Data mining the past is easy but its conclusions are specific to the specific period of backtesting and are not accurate forecasts of what is going to happen in the future if a different macroeconomic/market background occurs. Much of the period of 1972 - 2017 was spent in the greatest bull market in bonds in US history, a period when the Sharpe ratio of a bond portfolio was very high, great returns, no risk. There was a 30 year period during that time frame, 1982 -2011 when US 30 year bonds actually outperformed US stocks. That had not occurred since the 1800s to give you and idea of how unusual it was. Therefore bond heavy portfolios (70%) suffered little in the way of reduced return while vastly reducing portfolio risk.

We're in a different world now. Current estimates of the ten year Treasury returns are about what their current interest rate is, 2.7%. Inflation over the last 12 months has been 2.2% which produces a real inflation adjusted yield of 0.5% currently for 10 year Treasuries. Both rates and inflation are expected to remain fairly stable for the foreseeable future. If you want to put 70% of your portfolio in an asset with an expected 10 year real return of 0.5% you certainly can do so. It will undoubtedly vastly dampen down volatility. Your risk, in contrast of past decades, is that your overall portfolio returns are going to be much lower than in past decades when bonds rode the 34 year principal appreciation wave of ever lower rates (15% to less than 2%) and increasing real returns on a fixed yield due to ever decreasing inflation (12+% to less than 2%). Inflation is expected to remain in the 2.3% range going forward and bond yields are also not expected to rise significantly. The results of 1972 - 2017 period for 30 bonds/ 70 2X leverage is incredibly unlikely to repeat. Backtesting data mining is easy and fun but it does not accurately define the future. This is especially so for complex 30 equity/70 bond portfolios promising 60/40 returns with reduced volatility going forward. Complex portfolios if picked carefully in the rear view mirror may have shined brightly over the last 3 - 4 decades but IMO they are unlikely to do so going forward for the next 3 - 4 decades.

Garland Whizzer

User avatar
nisiprius
Advisory Board
Posts: 39463
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Portfolio of Leveraged Exchange Traded Funds

Post by nisiprius » Thu Jan 10, 2019 10:00 pm

long_gamma wrote:
Thu Jan 10, 2019 11:20 am
This analysis is flawed. You need to lever up both equity and bonds.
That is not what the authors say. They say
Downside risk is reduced using LETFs because the majority of the LETF portfolio is invested in a relatively safe bond fund.
They say nothing about using leverage on the bond side, too. If you do that, you are no longer talking about a "relatively safe bond fund."

They seem to be talking about something similar to the "Larry Portfolio," in which the stocks in a traditional portfolio are replaced with 100% small-cap value, the stock allocation is cut down, and the bond allocation is increased, so as to match standard deviation with the traditional portfolio. They seem to be presenting the same basic idea, using a leveraged stock ETF instead of a straight small-cap value stock ETF.

You seem to be talking about risk parity, but using only stocks and bonds.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

long_gamma
Posts: 352
Joined: Mon Sep 14, 2015 4:13 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by long_gamma » Thu Jan 10, 2019 10:31 pm

nisiprius wrote:
Thu Jan 10, 2019 10:00 pm
long_gamma wrote:
Thu Jan 10, 2019 11:20 am
This analysis is flawed. You need to lever up both equity and bonds.
That is not what the authors say. They say
Downside risk is reduced using LETFs because the majority of the LETF portfolio is invested in a relatively safe bond fund.
They say nothing about using leverage on the bond side, too. If you do that, you are no longer talking about a "relatively safe bond fund."

They seem to be talking about something similar to the "Larry Portfolio," in which the stocks in a traditional portfolio are replaced with 100% small-cap value, the stock allocation is cut down, and the bond allocation is increased, so as to match standard deviation with the traditional portfolio. They seem to be presenting the same basic idea, using a leveraged stock ETF instead of a straight small-cap value stock ETF.

You seem to be talking about risk parity, but using only stocks and bonds.
I didn't read the paper completely. It is not ground breaking research, since the traders used this method for eons.

I looked at the the Table, they are levering up bonds. In the table 2, the are levering up the 7 to 10 year T-bonds and 20 year T-bonds by 2 or 3 times.

Relatively safe bond is the cash portion of the levered portfolio, in this case it is Bond ladders.

Image
"Everyone has a plan 'till they get punched in the mouth." --Mike Tyson

User avatar
Random Musings
Posts: 5551
Joined: Thu Feb 22, 2007 4:24 pm
Location: Pennsylvania

Re: Portfolio of Leveraged Exchange Traded Funds

Post by Random Musings » Thu Jan 10, 2019 11:04 pm

garlandwhizzer wrote:
Thu Jan 10, 2019 12:33 pm
It is a very simple exercise in mathematics over a given time frame to create a complex portfolio using leverage/alternates/options, etc., with higher bond allocations that outperforms a standard 60/40 portfolio and provides a higher Sharpe ratio. Data mining the past is easy but its conclusions are specific to the specific period of backtesting and are not accurate forecasts of what is going to happen in the future if a different macroeconomic/market background occurs. Much of the period of 1972 - 2017 was spent in the greatest bull market in bonds in US history, a period when the Sharpe ratio of a bond portfolio was very high, great returns, no risk. There was a 30 year period during that time frame, 1982 -2011 when US 30 year bonds actually outperformed US stocks. That had not occurred since the 1800s to give you and idea of how unusual it was. Therefore bond heavy portfolios (70%) suffered little in the way of reduced return while vastly reducing portfolio risk.

We're in a different world now. Current estimates of the ten year Treasury returns are about what their current interest rate is, 2.7%. Inflation over the last 12 months has been 2.2% which produces a real inflation adjusted yield of 0.5% currently for 10 year Treasuries. Both rates and inflation are expected to remain fairly stable for the foreseeable future. If you want to put 70% of your portfolio in an asset with an expected 10 year real return of 0.5% you certainly can do so. It will undoubtedly vastly dampen down volatility. Your risk, in contrast of past decades, is that your overall portfolio returns are going to be much lower than in past decades when bonds rode the 34 year principal appreciation wave of ever lower rates (15% to less than 2%) and increasing real returns on a fixed yield due to ever decreasing inflation (12+% to less than 2%). Inflation is expected to remain in the 2.3% range going forward and bond yields are also not expected to rise significantly. The results of 1972 - 2017 period for 30 bonds/ 70 2X leverage is incredibly unlikely to repeat. Backtesting data mining is easy and fun but it does not accurately define the future. This is especially so for complex 30 equity/70 bond portfolios promising 60/40 returns with reduced volatility going forward. Complex portfolios if picked carefully in the rear view mirror may have shined brightly over the last 3 - 4 decades but IMO they are unlikely to do so going forward for the next 3 - 4 decades.

Garland Whizzer
I concur. My grandfather retired in 1980, and being "conservative", put most of his retirement monies into LT bonds and some in equities. Where you stand in time can really drive investment results; SWR was really no issue for him. Today, 3% bonds are not going to generate a great deal in real returns unless we have a long period of deflation. But I don't think anyone (well, very few) wants to see negative bond yields as seen in a few countries around the world today.

I believe the portfolio analyzer tool can send people down the wrong path. Past results are just that, past results.

RM
I figure the odds be fifty-fifty I just might have something to say. FZ

AlphaLess
Posts: 1607
Joined: Fri Sep 29, 2017 11:38 pm
Location: Kentucky

Re: Portfolio of Leveraged Exchange Traded Funds

Post by AlphaLess » Thu Jan 10, 2019 11:18 pm

I read all the comments. Very interesting and thought-provoking discussions.

Seems like there are some disagreements among the forum readers on the following:
- what authors say vs what they actual do,
- what authors do vs what they *SHOULD* do,
- whatever authors do is "xyz".

Those considerations aside, how much does the success of the LEVERAGED vs STANDARD strategy depends on the following:
- the cost of leverage: the steepness of the yield curve, or the additional yield over various products (stock or bond) over short-term bonds,
- the cost of leveraged products (in terms of expense ratios and slippage),
- unique to the period studied.

Does anyone have a short mathematical description that presents a STYLIZED model of this, with some basic assumptions, including:
- assumptions about the basic (non-leverage) asset classes, most importantly, expected return: E and covariane matrix, C.
- assumptions about leveraged products: manner of rebalancing, cost of leverage, expense ratios,
- additional assumptions about the steepness of the yield curve,
- the considerations.

The model could provide a solution in terms of portfolio weights to apply to leveraged products, given an input of portfolio weights to the basic products.

Seems like this *SHOUL* work under certain conditions, and *MAY NOT* work under other conditions.
"A Republic, if you can keep it". Benjamin Franklin. 1787. | Party affiliation: Vanguard. Religion: low-cost investing.

AlphaLess
Posts: 1607
Joined: Fri Sep 29, 2017 11:38 pm
Location: Kentucky

Re: Portfolio of Leveraged Exchange Traded Funds

Post by AlphaLess » Fri Jan 11, 2019 12:08 am

long_gamma wrote:
Thu Jan 10, 2019 10:31 pm

Image
Thanks for sharing this.

When expressed in this terms, let's consider differences between these portfolios:
- non-leveraged,
- LEV-2,
- LEV-3

Let's net out on a row-by-row basis. Let's also assume an idealistic situation in which we re-balance daily. In practice, we won't be doing this, but for this exercise, let's just consider it.

For example, the difference between the non-leveraged and LEV-2 portfolios, considering row-by-row:
- for large caps:
-- we have 50% exposure in non-lev, at no carrying cost,
-- we have 25% exposure to 2X lev, at carrying cost equal to: short term funding plus expense ratio.

The same is true for every one of the lines: mid-cap, small-cap, 20Y T-bond, and 7-10 T Bonds.

In effect, we are leveraging the non-leveraged portfolio 2x, at HALF the weight, thus getting exact same exposure.
Our cost is short-term funding fee plus expense ratio (which will be reflected in the leveraged product underperformance).

Notice that our funding cost is over the leveraged part of our portfolio. For the LEV-2, 50% of the portfolio has 2x leveraged, thus we are carrying funding and expense ratios on the 50% of our GMV.

But we are buying 50% 0-7 year bond ladder with the 'freed-up. capital.

THUS, in effect, we are borrowing money, and buying 0-7 bond ladder, and paying expense ratios.


So the ENTIRE outperformance of the leveraged strategy depends on:
- yield curve steepness from 0 to 7 years,
- expense ratios of leveraged funds.

If we knew that 0-7 year bond curve will remain above the short-term borrowing cost (0-term borrowing cost), and we can borrow at the same rate as top qualify financial institutions, then sure, this strategy MAY work.

Now, notice that the whole "borrow 0-term, and buy a bond ladder" strategy is a side-kick to the MAIN, unleveraged stock+bond porfolio.

So our risk (standard deviation, daily fluctuations, etc) are for the most part the same as for the unleveraged portfolio.

It is due to the fact that:
- "the borrow short to buy a bond ladder" strategy is profitable, AND
- that the above strategy is some uncorrelated, or maybe even negatively correlated to the main, unleveraged portfolio that we experience higher return and higher sharp.

But fundamentally, this leveraged strategy may not work, if:
- the "juice" from borrowing short and buying a bond ladder does not have enough juice (SUCH AS NOW), AND
- the expenses kill the goose.

According to this page:
https://www.treasury.gov/resource-cente ... data=yield

1M yield: 2.40%
7-year yield: 2.60%.

Average juice: (2.6 - 2.4)% / 2 = 0.1%.

That's not enough juice for the leveraged strategy to work, especially given the fees.
"A Republic, if you can keep it". Benjamin Franklin. 1787. | Party affiliation: Vanguard. Religion: low-cost investing.

long_gamma
Posts: 352
Joined: Mon Sep 14, 2015 4:13 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by long_gamma » Fri Jan 11, 2019 2:46 am

AlphaLess wrote:
Fri Jan 11, 2019 12:08 am

That's not enough juice for the leveraged strategy to work, especially given the fees.
Expense ratio was taken into consideration, since my portfolio charts from 2007 and other poster from 1998 were real life funds.

Agree overall with premise of your post. Along with the flat yield curve, increase in volatility will kill any out-performance. Current market is not very favorable for the leverage strategy. Sequence of risk is another risk for these type of strategy. While some of the froth in valuation has come down, both bonds and stocks going down simultaneously because of rising rate will kill the goose egg.

I was just replying to following hyperbole statements by Nisipirius. It was certainly not the case.
nisiprius wrote:
Thu Jan 10, 2019 9:45 am

in real life, no advantage would have been seen in using a leveraged fund portfolio (red line) over a conventional fund portfolio.


no reduction in downside risk would have been seen in in the leveraged portfolio, either in 2008-2009 or in 2000-2002.
"Everyone has a plan 'till they get punched in the mouth." --Mike Tyson

long_gamma
Posts: 352
Joined: Mon Sep 14, 2015 4:13 pm

Re: Portfolio of Leveraged Exchange Traded Funds

Post by long_gamma » Fri Jan 11, 2019 3:57 am

AlphaLess wrote:
Fri Jan 11, 2019 12:08 am

So the ENTIRE outperformance of the leveraged strategy depends on:
- yield curve steepness from 0 to 7 years,
- expense ratios of leveraged funds.

If we knew that 0-7 year bond curve will remain above the short-term borrowing cost (0-term borrowing cost), and we can borrow at the same rate as top qualify financial institutions, then sure, this strategy MAY work.

Now, notice that the whole "borrow 0-term, and buy a bond ladder" strategy is a side-kick to the MAIN, unleveraged stock+bond porfolio.

So our risk (standard deviation, daily fluctuations, etc) are for the most part the same as for the unleveraged portfolio.

It is due to the fact that:
- "the borrow short to buy a bond ladder" strategy is profitable, AND
- that the above strategy is some uncorrelated, or maybe even negatively correlated to the main, unleveraged portfolio that we experience higher return and higher sharp.

But fundamentally, this leveraged strategy may not work, if:
- the "juice" from borrowing short and buying a bond ladder does not have enough juice (SUCH AS NOW), AND
- the expenses kill the goose.

According to this page:
https://www.treasury.gov/resource-cente ... data=yield

1M yield: 2.40%
7-year yield: 2.60%.

Average juice: (2.6 - 2.4)% / 2 = 0.1%.

That's not enough juice for the leveraged strategy to work, especially given the fees.
To add further, we can look at two portfolio's

Portfolio1:
60/40 VFINX/VBMFX

Portfolio2:
Similar to Efficientinvest portfolio
ULPIX (Stocks 2xbull): 30%, WHOSX: Treasury (Extended duration ) 33%, I will use short term investment grade VFSTX to control the standard deviation and bring it to par with Portfolio1. Average duration is shorter than the bond ladder duration used in the paper 2.5 vs 4,175 years, but makes up with credit risk. VFSTX 37%
https://www.portfoliovisualizer.com/ba ... tion5_1=40

Image


So both portfolio has similar Std. Devn. Portfolio2 certainly has better return and lower draw down. But most of the out-performanance came after Nov 2008. Let us look at volatility and yield curve chart.

Image

Red line indicates end of Oct 2008. Yield difference was pretty high during Nov 2008 and Volatility was at its peak during that time. So most of the out performance was because of crushing vola and paid handsomely for the yield ride down.

One thing attractive was that drawdown was much lower during crisis in levered portfolio. Another thing to note is that portfolio formation date 1998, yield curve and vola was similar to current date.
"Everyone has a plan 'till they get punched in the mouth." --Mike Tyson

User avatar
LadyGeek
Site Admin
Posts: 57800
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Portfolio of Leveraged Exchange Traded Funds

Post by LadyGeek » Fri Jan 11, 2019 9:25 am

I removed several off-topic posts. As a reminder, see: General Etiquette
We expect this forum to be a place where people can feel comfortable asking questions and where debates and discussions are conducted in civil tones.

At all times we must conduct ourselves in a respectful manner to other posters. Attacks on individuals, insults, name calling, trolling, baiting or other attempts to sow dissension are not acceptable.
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

AlphaLess
Posts: 1607
Joined: Fri Sep 29, 2017 11:38 pm
Location: Kentucky

Re: Portfolio of Leveraged Exchange Traded Funds

Post by AlphaLess » Fri Jan 11, 2019 10:39 am

long_gamma wrote:
Fri Jan 11, 2019 2:46 am
AlphaLess wrote:
Fri Jan 11, 2019 12:08 am

That's not enough juice for the leveraged strategy to work, especially given the fees.
Expense ratio was taken into consideration, since my portfolio charts from 2007 and other poster from 1998 were real life funds.

Agree overall with premise of your post. Along with the flat yield curve, increase in volatility will kill any out-performance. Current market is not very favorable for the leverage strategy. Sequence of risk is another risk for these type of strategy. While some of the froth in valuation has come down, both bonds and stocks going down simultaneously because of rising rate will kill the goose egg.

I was just replying to following hyperbole statements by Nisipirius. It was certainly not the case.
nisiprius wrote:
Thu Jan 10, 2019 9:45 am

in real life, no advantage would have been seen in using a leveraged fund portfolio (red line) over a conventional fund portfolio.


no reduction in downside risk would have been seen in in the leveraged portfolio, either in 2008-2009 or in 2000-2002.
Thanks for the informative comments.

Two thoughts:
- whenever analyzing a strategy, decomposing it to 'factors' (intentionally vague) is important because otherwise you don't know why it works,
- I do agree that in history (the last 15-25 years), it actually worked.

So, can we decompose it even further:
- e.g., what exposure are we taking by rebalancing daily, monthly, or quarterly?
- are we missing anything?

If I am not mistaken, leveraged funds are long gamma, right (no pun intended with respect to your username)?

Thus, an UP-UP move would reward the leveraged strategy more than the non-leveraged one.

A daily rebalance of the leveraged part of the portfolio would get rid of it.
But a daily rebalance is not practical.

So, by rebalancing monthly, what type of exposure are we taking?
Is it possible to quantify it?

A simple way to quantify it would be:
- measure volatility on daily returns, say, it is sd_d,
- measure volatility on monthly returns, say, it is sd_m,
- and if daily and monthly vols don't scale (sd_d == sd_m * sqrt(20.5)), then due to some 'momentum' exposure, the leveraged portfolio returns are different from the non-leveraged ones.

That last risk may not be worthwhile carrying.
"A Republic, if you can keep it". Benjamin Franklin. 1787. | Party affiliation: Vanguard. Religion: low-cost investing.

Post Reply